Machine learning forecasting of CR import from PRC in context of mutual PRC and USA sanctions

Institute of Technology and Business, School of Expertness and Valuation, Okružní 517/10, 37001 České Budějovice, Czech Republic

^* Corresponding author: machova@mail.vstecb.cz

Abstract

Mutual trade restrictions between the USA and the PRC caused by the USA feeling of imbalance of trade between these two countries have significantly influenced not only the trade between these two states but also the overall atmosphere of the international trade in the last few years. The objective of the contribution is to find out whether machine learning forecasting is capable of equalizing time series so that the model effectively forecasts the future development of the time series even in the context of an extraordinary situation caused by such factors as the mutual sanctions of the USA and PRC. The dataset shows the course of the time series at monthly intervals starting from January 2000 to June 2019. There is regression carried out using neural structures. Three sets of artificial neural networks are generated. They are differ in the considered time series lag. 10,000 neural networks are generated, out of which 5 with the best characteristics are retained. The mutual USA and PRC sanctions did not affect the success rate of the machine learning forecasting of the CR import from the PRC. It is evident that the mutual sanctions shall affect the trade between the CR and the PRC.